Agenda Request - threat model for first party data

[marianapr]

Several API designs rely on using first party data to obtain better utility. Since, the computation infrastructure is most often not ran by the party who owns such first party data, there is a question whether sharing such first party data with the parties running the measurement infrastructure is OK or whether we should aim to provide privacy for such data. The privacy issue here is not the usual user privacy we have discussed - there is a party who owns this data (however, this party may have commitments on how it uses/shares such data). The question here is whether we want to have designs that require sharing of first party data in order to obtain utility.

More discussion is this issue: patcg-individual-drafts/ipa#77

20 min talk + 25 min discussion

[AramZS]

On review, I'm 90% sure the discussion portion will expand beyond the designated time here and I'll reserve at least an hour

[AramZS]

@marianapr Would you like to follow up on this in the upcoming meeting?

[marianapr]

I am not sure there is any need to follow - I think based on the discussion we made a decision and I will be writing a paragraph that reflects the decisions to be added to the threat model document.